{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LG3SKJYAO3GXR5UBLUZS5MWUHC","short_pith_number":"pith:LG3SKJYA","schema_version":"1.0","canonical_sha256":"59b725270076cd78f6815d332eb2d438a727f5cbde83206542083f7bbfa24bd8","source":{"kind":"arxiv","id":"2508.12566","version":1},"attestation_state":"computed","paper":{"title":"Help or Hurdle? Rethinking Model Context Protocol-Augmented Large Language Models","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Haonan Zhong, Jingling Xue, Wei Song, Yuekang Li, Ziqi Ding","submitted_at":"2025-08-18T02:06:05Z","abstract_excerpt":"The Model Context Protocol (MCP) enables large language models (LLMs) to access external resources on demand. While commonly assumed to enhance performance, how LLMs actually leverage this capability remains poorly understood. We introduce MCPGAUGE, the first comprehensive evaluation framework for probing LLM-MCP interactions along four key dimensions: proactivity (self-initiated tool use), compliance (adherence to tool-use instructions), effectiveness (task performance post-integration), and overhead (computational cost incurred). MCPGAUGE comprises a 160-prompt suite and 25 datasets spanning"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2508.12566","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/publicdomain/zero/1.0/","primary_cat":"cs.AI","submitted_at":"2025-08-18T02:06:05Z","cross_cats_sorted":[],"title_canon_sha256":"7552a09e224b587fbbfac10291cea1ff09e381dd78f9ce1870054e212c472fa8","abstract_canon_sha256":"f512dafa201ccf1ac72306a5d2f5b96696fde90e10fa851e4625410e38b1df0f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:55:23.261001Z","signature_b64":"xVuKwzV/XQqyuqZSX5u9QEY5pzhkgJGO/FdWtlR7sE0OjNVXf/gdn02SpTn4hdr9SiUW359IT0oePolXfAdkDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"59b725270076cd78f6815d332eb2d438a727f5cbde83206542083f7bbfa24bd8","last_reissued_at":"2026-07-05T11:55:23.260627Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:55:23.260627Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Help or Hurdle? Rethinking Model Context Protocol-Augmented Large Language Models","license":"http://creativecommons.org/publicdomain/zero/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Haonan Zhong, Jingling Xue, Wei Song, Yuekang Li, Ziqi Ding","submitted_at":"2025-08-18T02:06:05Z","abstract_excerpt":"The Model Context Protocol (MCP) enables large language models (LLMs) to access external resources on demand. While commonly assumed to enhance performance, how LLMs actually leverage this capability remains poorly understood. We introduce MCPGAUGE, the first comprehensive evaluation framework for probing LLM-MCP interactions along four key dimensions: proactivity (self-initiated tool use), compliance (adherence to tool-use instructions), effectiveness (task performance post-integration), and overhead (computational cost incurred). MCPGAUGE comprises a 160-prompt suite and 25 datasets spanning"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.12566","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2508.12566/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2508.12566","created_at":"2026-07-05T11:55:23.260684+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.12566v1","created_at":"2026-07-05T11:55:23.260684+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.12566","created_at":"2026-07-05T11:55:23.260684+00:00"},{"alias_kind":"pith_short_12","alias_value":"LG3SKJYAO3GX","created_at":"2026-07-05T11:55:23.260684+00:00"},{"alias_kind":"pith_short_16","alias_value":"LG3SKJYAO3GXR5UB","created_at":"2026-07-05T11:55:23.260684+00:00"},{"alias_kind":"pith_short_8","alias_value":"LG3SKJYA","created_at":"2026-07-05T11:55:23.260684+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.07461","citing_title":"Mitigating Taint-Style Vulnerabilities in MCP Servers via Security-Aware Tool Descriptions","ref_index":51,"is_internal_anchor":true},{"citing_arxiv_id":"2604.17234","citing_title":"From Language to Action: Enhancing LLM Task Efficiency with Task-Aware MCP Server Recommendation","ref_index":72,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC","json":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC.json","graph_json":"https://pith.science/api/pith-number/LG3SKJYAO3GXR5UBLUZS5MWUHC/graph.json","events_json":"https://pith.science/api/pith-number/LG3SKJYAO3GXR5UBLUZS5MWUHC/events.json","paper":"https://pith.science/paper/LG3SKJYA"},"agent_actions":{"view_html":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC","download_json":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC.json","view_paper":"https://pith.science/paper/LG3SKJYA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.12566&json=true","fetch_graph":"https://pith.science/api/pith-number/LG3SKJYAO3GXR5UBLUZS5MWUHC/graph.json","fetch_events":"https://pith.science/api/pith-number/LG3SKJYAO3GXR5UBLUZS5MWUHC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC/action/storage_attestation","attest_author":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC/action/author_attestation","sign_citation":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC/action/citation_signature","submit_replication":"https://pith.science/pith/LG3SKJYAO3GXR5UBLUZS5MWUHC/action/replication_record"}},"created_at":"2026-07-05T11:55:23.260684+00:00","updated_at":"2026-07-05T11:55:23.260684+00:00"}